Neural‐Network‐Based Event‐Triggered Consensus Tracking for Nonlinear Heterogeneous Multi‐Agent Systems Against Denial‐of‐Service Attacks

This article proposes a distributed resilient control framework to ensure consensus tracking for nonlinear heterogeneous multi‐agent systems (NHMAS) under aperiodic denial‐of‐service (DoS) attacks. First, a resilient observer is introduced to estimate the leader's state considering that not all agents have direct access to the leader's information. This observer incorporates a local asynchronous dynamic event‐triggered mechanism that effectively mitigates the impact of DoS attacks. Second, an event‐driven estimator is designed to reconstruct the unmeasured states. To address unknown nonlinear dynamics, an adaptive output‐feedback compensator based on neural networks is presented and only updated at event‐triggered instants. Finally, a distributed control framework is developed to achieve consensus tracking for NHMAS in the presence of DoS attacks. Theoretical analysis demonstrates that the closed‐loop system guarantees resilient consensus despite DoS attacks, and Zeno behavior is rigorously excluded by ensuring a strictly positive minimum inter‐event time. Comparative simulations validate that the proposed framework effectively enhances system resilience, accelerates convergence, and reduces computational overhead. Experiments on a group of unmanned ground vehicles further verify the advantages of the proposed method.